The Gaia-ESO Survey: the origin and evolution of s-process elements · 2018. 6. 11. · and Zr, and three belonging to the second peak, Ba, La and Ce - using the Gaia-ESO idr5 results

Astronomy & Astrophysics manuscript no. magrini_neutron c©ESO 2018June 11, 2018

The Gaia-ESO Survey: the origin and evolution of s-processelements ?

L. Magrini1, L.Spina2, S. Randich1, E. Friel,3, G. Kordopatis4, C. Worley5, E. Pancino1, 6, A. Bragaglia7, P. Donati7, G.Tautvaišiene8, V. Bagdonas8, E. Delgado-Mena9, V. Adibekyan9, S. G. Sousa9, F. M. Jiménez-Esteban10, N. Sanna1, V.

Roccatagliata1, R. Bonito11, L. Sbordone12, S. Duffau13, G. Gilmore5, S. Feltzing14, R. D. Jeffries15, A. Vallenari16,E. J. Alfaro17. T. Bensby14, P. Francois18, S. Koposov5, A. J. Korn19, A. Recio-Blanco20, R. Smiljanic21, A. Bayo22, 23,G. Carraro24, A. R. Casey2, M. T. Costado25, F. Damiani11, E. Franciosini1, A. Frasca26, A. Hourihane5, P. Jofré27, P.

de Laverny20, J. Lewis5, T. Masseron28, 29, L. Monaco13, L. Morbidelli1, L. Prisinzano11, G. Sacco1, S. Zaggia16

(Affiliations can be found after the references)

Received ; accepted

ABSTRACT

Context. Several works have found an increase of the abundances of the s-process neutron-capture elements in the youngest Galacticstellar populations. These trends give important constraints to stellar and Galactic evolution and they need to be confirmed with largeand statistically significant samples of stars spanning wide age and distance intervals.Aims. We aim to trace the abundance patterns and the time-evolution of five s-process elements - two belonging to the first peak, Yand Zr, and three belonging to the second peak, Ba, La and Ce - using the Gaia-ESO idr5 results for open clusters and disc stars.Methods. From the UVES spectra of cluster member stars, we determined the average composition of clusters with ages >0.1 Gyr.We derived statistical ages and distances of field stars, and we separated them in thin and thick disc populations. We studied the time-evolution and dependence on metallicity of abundance ratios using open clusters and field stars whose parameters and abundanceswere derived in a homogeneous way.Results. Using our large and homogeneous sample of open clusters, thin and thick disc stars, spanning an age range larger than 10 Gyr,we confirm an increase towards young ages of s-process abundances in the Solar neighbourhood. These trends are well defined foropen clusters and stars located nearby the solar position and they may be explained by a late enrichment due to significant contributionto the production of these elements from long-living low-mass stars. At the same time, we found a strong dependence of the s-processabundance ratios with the Galactocentric distance and with the metallicity of the clusters and field stars.Conclusions. Our results, derived from the largest and homogeneous sample of s-process abundances in the literature, confirm thegrowth with decreasing stellar ages of the s-process abundances in both field and open cluster stars. At the same time, taking advantageof the abundances of open clusters located in a wide Galactocentric range, they open a new view on the dependence of the s-processevolution on the metallicity and star formation history, pointing to different behaviours at various Galactocentric distances.

Key words. Galaxy: abundances, open clusters and associations: general, open clusters and associations: individual: Berkeley 25,Berkeley 31, Berkeley 36, Berkeley 44, Berkeley 81, Pismis 18, Trumpler 20, Trumpler 23, M67, NGC 4815, NGC 6705, NGC 6802,NGC 6005, NGC 2516, NGC 6633, NGC 2243, NGC 6067, NGC 6259, NGC 3532, NGC 2420, Melotte 71, Rup 134, Galaxy: disc

1. Introduction

Elemental abundances and abundance ratios in the different pop-ulations of our Galaxy provide fundamental constraints to thescenarios of galaxy formation and evolution. With a large num-ber of different elements we can study the different processesand time scales involved in stellar and galactic evolution.

The heavy elements (atomic number Z>30) are produced bysubsequent capture of neutrons on lighter elements. Their evolu-tion in the Galaxy has been studied since the pioneering work ofPagel & Tautvaisiene (1997); Travaglio et al. (1999). There are54 stable or long-lived neutron-capture elements, compared toonly 30 lighter elements (cf. Sneden et al. 2008). However, theyare much less abundant than the lighter elements and correspondto an abundance by number of 10−8 in the Sun. Their production

? Based on observations collected with the FLAMES instrument atVLT/UT2 telescope (Paranal Observatory, ESO, Chile), for the Gaia-ESO Large Public Spectroscopic Survey (188.B-3002, 193.B-0936).

is an endothermic process. In addition, the Coulomb barriers in-crease with the proton number. Consequently, the nuclei heavierthan iron can be created only with a series of neutron captureevents: for those events there are no Coulomb barriers. Neutron-capture elements can be split into two different groups: thoseformed mainly via slow neutron-capture “s-process” and thoseproduced by rapid neutron capture “r-process”. The terms slow(time-scale for a single neutron capture of hundreds or thousandsof years) or rapid (fractions of second) are so defined in compar-ison with the time-scales of the β-decay of the nuclei on whichneutrons are accreted. For the s-process and r-process, low, i.e.107 to 1011 cm−2 s−1, and high fluxes of neutrons, i.e. ∼1020 cm−2

s−1, are needed, respectively.

The s-process dominated elements, hereafter s-process ele-ments, are distributed primarily in two peaks in the periodic ta-ble around the neutron magic numbers N = 50, 82. The first s-process-peak is located around N=50 and leads to the formationof Sr, Y, and Zr (light-s elements); the second peak at N=82 pro-

Article number, page 1 of 16

arX

iv:1

806.

0306

8v1

[as

tro-

ph.G

A]

8 J

un 2

018

A&A proofs: manuscript no. magrini_neutron

duces Ba, La, Ce, Pr, and Nd. There is also a third peak that cor-responds to the production of Pb at N=126. The s-process canbe further divided into a “main” process, a “weak” process anda “strong" process: the first occurs during the asymptotic giantbranch (AGB) phases (Gallino et al. 1998; Busso et al. 1999),whereas the second takes place in massive stars that produce el-ements with N≤88 (Raiteri et al. 1993; Pignatari et al. 2010) andthe third is responsible for about 50% of solar 208Pb productionby low-metallicity AGB stars (e.g. Gallino et al. 1998).

Recently, several works found an increase of the abundancesof the s-process elements for the youngest stellar populations.Among the first studies that noticed it, D’Orazi et al. (2009) re-ported the discovery of increasing barium abundance with de-creasing age for a large sample of Galactic open clusters. Theygave a first tentative explanation assuming a higher Ba yieldfrom the s-process in low-mass stars than the average suggestedby parameterised models of neutron-capture nucleosynthesis.However, their chemical evolution model was not able to explainthe very high [Ba/Fe] measured in the youngest open clusters oftheir sample. Other studies followed, all finding strong trends ofBa abundance with age.

A number of studies (e.g., Maiorca et al. 2011; Jacobson &Friel 2013; Mishenina et al. 2013, 2015) extended the analysisto other s-process elements, both lighter (Y, Zr) and heavier (Ba,La, Ce) neutron-rich nuclei. Conclusions varied with the studies,painting a somewhat confused picture of age dependencies in el-ements other than Ba. Maiorca et al. (2011) found the elementsY, Ce, Zr, and La to show increasing abundance ratios with de-creasing age, as with Ba, but to a lesser degree. Jacobson & Friel(2013) found strong age trends in Ba but no age trends in Zrand La, and noted that the conclusions drawn by different stud-ies depended on the sample size and the range of ages sampled,and were affected by sometimes large systematic differences be-tween studies (e.g. Yong et al. 2012). Mishenina et al. (2013,2015) found clusters and field stars shared similar behavior in Yand La, with slight, if any, trends with age. A number of studiesfound overall enhancements in abundance ratios in the clustersrelative to solar, up to 0.2 dex, particularly for [Zr/Fe].

Clusters in these samples not only spanned large age ranges,from several hundred million years to ∼8 Gyr, they also cov-ered large distance ranges, in particular sampling the outer disk.Although open clusters have been shown not to follow any rela-tionship between age and metallicity, they do exhibit gradientsin abundance, and their location (at birth) is more of a factor indetermining their abundance than is their age. The interplay ofthe effects of both age and location introduce complexity into theinterpretation of abundance trends, while at the same time offer-ing the potential to illuminate them. While these cluster studiesbenefitted from the accurate age determinations of open clus-ters, particularly their ability to trace the recent epoch, they nev-ertheless revealed that differing sample sizes, age, and distancedistributions, and the non-uniformity of analysis between stud-ies complicated the conclusions that could be drawn, particularlyabout weak trends.

Other groups investigated this phenomenon using field stars,especially solar twins. We recall that the ages of field stars canbe determined with lower accuracy than those for open clustersat least in the pre-Gaia era. The sample of solar-type stars of daSilva et al. (2012) showed a pattern of [s/Fe] with age similar tothat of open clusters (D’Orazi et al. 2009; Maiorca et al. 2012),with increasing abundance ratios with decreasing age, althoughthe form of the dependence varied with element: for Ce and Zr,the increase pertained only to stars younger than the Sun, whilefor Y and Sr, the abundances increased linearly with decreasing

age. Recently, Reddy & Lambert (2017) measured La, Ce, Nd,and Sm: their Figure 7 suggests that the [X/Fe] values measuredfor these heavy s-process elements display a smoothly increasingweak trend with decreasing stellar age, with changes of ∼0.1 dexfrom 10 Gyr ago to the present. For Ba, they found, like manyothers, a much stronger age dependence. They explain the largerenhancement for Ba in younger stars as due to an overestimationof Ba by standard methods of LTE analysis, finding the Ba en-hancements to be strongly correlated with the level of chromo-spheric activity. If this effect is considered and the very youngstars are excluded (the chromospheric activity has a striking de-cline from 1-10 Myr to the age of the Sun), the enhancementof Ba is in line with that of the other s-process elements. In thesame framework, Nissen et al. (2017) analysed a sample of tenstars from the Kepler LEGACY (Silva Aguirre et al. 2017) andfound that over the lifetime of the Galactic disc the abundance ofY increases. The trend of [Y/Fe] with age, and an even strongertrend of [Y/Mg], were previously found by e.g., da Silva et al.(2012); Nissen (2016); Spina et al. (2016); Feltzing et al. (2017)and it can be explained by production of yttrium via the slowneutron-capture process in low-mass AGB stars (Maiorca et al.2012; Karakas & Lugaro 2016). Finally, Spina et al. (2018), intheir differential analysis of a sample of solar twin stars, foundan increase with decreasing stellar ages of the abundance ratios[X/Fe] of all the s-process elements of the first and second peaks.They found that the rise of Ba with time is consistent with that ofthe other n-capture elements, considering its higher contributionfrom the s-process. In all cases, trends of [X/Fe] are weak, withchanges of 0.1 to 0.2 dex over the lifetime of the disk.

Thus, a general consensus of an increasing, though weak,trend in the s- process elements production exists. There are stillissues regarding the maximum abundance for Ba, which is stilldebated (see Reddy & Lambert 2017) and the trends for the openclusters, which are affected by positional dependence, radial gra-dients, and differences in analysis and samples (see, e.g. Jacob-son et al. 2011; Yong et al. 2012). Taking inspiration from theconclusions of Jacobson & Friel (2013), we need a homogeneousanalysis of large samples of open clusters together with a sam-ple of field stars that will facilitate the measurement of many n-capture species in clusters and in the field spanning a wide rangein age.

Among the several on-going spectroscopic surveys, theGaia-ESO Survey (GES, Gilmore et al. 2012; Randich et al.2013), a ESO large public survey, is providing high resolu-tion spectra of different stellar populations of our Galaxy usingFLAMES@VLT (Pasquini et al. 2002) employing both Giraffeand UVES fibres in Medusa mode. It aims at homogeneouslydetermining stellar parameters and abundances for a large sam-ple of stars in the field, in Galactic open clusters, and in cali-bration samples, including also globular clusters. In particular,the high resolution spectra obtained with UVES allow the deter-mination of abundances of a large variety of elements includingmany neutron-capture elements: two light s-process elements (Y,Zr), three heavy s-process elements (Ba, La, Ce), one r-processelement (Eu), and three elements with significant contributionsfrom both processes (mixed elements– Mo, Pr, and Nd). In thepresent work we investigate the evolution of the group of s-process elements. The paper is structured as follows. In Section2 we present the data reduction and analysis and in Section 3 thesolar abundance scale. In Sections 4 we describe the field andcluster samples, whereas in Section 5 we show the results on theabundance of s-process elements, and in Section 6 we discusstheir time evolution. Finally, in Section 7 we give our summaryand conclusions.


Magrini, L. et al.: S-process in Open Clusters and Galactic discs

2. The data reduction and analysis

We use data from the 5th internal data release (idr5) of the Gaia-ESO Survey. For this work we use only the products bases onthe UVES spectra, which have the higher spectral resolution(R=47000) and wider spectral range. They have been reducedand analysed by the Gaia-ESO consortium. UVES data reductionis carried out using the FLAMES-UVES ESO public pipeline(Modigliani et al. 2004). The UVES data reduction process andthe determination of the radial velocities (RVs) are describedin Sacco et al. (2014). Various WGs contribute to the spectralanalysis of different kinds of stars and/or setups: the data dis-cussed in the present paper have been analysed by WG11 whichis in charge of the analysis of the UVES spectra of F-G-K spec-tral type stars both in the field of the Milky Way (MW) and inintermediate-age and old clusters and obtained with the UVESred arm standard setting with central wavelength of 520 and 580nm. These spectra were analysed with the Gaia-ESO multiplepipelines strategy, as described in Smiljanic et al. (2014). Theresults of each pipeline are combined with an updated method-ology (Casey et al., in preparation) to define a final set of rec-ommended values of the atmospheric parameters. The results ofWG11 are homogenised using several calibrators e.g., bench-mark stars and open/globular clusters selected as described inPancino et al. (2017) and adopted for the homogenisation byWG15 (Hourihane et al. in preparation). The recommended pa-rameters and abundances used in the present work are reported inthe final idr5 catalog, which contains the observations obtainedbefore December 2015.

To measure the usually faint absorption lines of neutron cap-ture elements a high signal-to-noise ratio (SNR) moth spectrumis required. For this reason, for the field stars we made a selec-tion of the UVES results based on the SNR and on the qual-ity of the parameter determination. The thresholds of our cutsare defined comparing our results with the very high SNR sam-ples of Bensby et al. (2014) for Ba and of Battistini & Bensby(2016) for the other elements, designed to study the neutron-capture element abundances. To obtain similar dispersions asin Bensby et al. (2014) and in Battistini & Bensby (2016) forthe neutron-capture element distributions (at a given metallic-ity) we need a minimum SNR of 70 for Milky Way field stars.In addition, we discard stars whose errors on stellar parame-ters are: ETeff >150 K, Elogg >0.25 dex, E[Fe/H] >0.20 dex andEξ >0.20 km s−1. The final sample of field stars fulfilling theserequirements includes about 600 objects. The sample of openclusters includes 22 clusters, with a total of 165 member stars.

For the cluster stars, we did not apply any selection, becausetheir spectra usually are characterised by higher SNR (and thusalmost all stars in cluster stars will be included). In addition,we do not consider individual measurements for cluster stars,but we rely on the cluster median abundances, which are derivedfrom several member stars (the membership analysis is discussedin Section 4.2). We only discard cluster abundances with highstandard deviations > 0.3 dex.

3. Solar abundance scale

To obtain abundances on the solar scale, we need to define ourabundance reference. In Table 1 we show three different setsof abundances: the solar abundances from idr5 computed fromarchive solar spectra; those from Grevesse et al. (2007); and theabundances of giant stars in M67. The abundances are in theusual 12+log(X/H) form. The cluster M67 is known to have thesame composition as the Sun (e.g. Randich et al. 2006; Pasquini

et al. 2008; Önehag et al. 2014; Liu et al. 2016) and thus it isuseful to confirm it with the idr5 data. Since small effects ofstellar diffusion might be present in the abundances of the dwarfstars of M67 (see, e.g., Önehag et al. 2014 and Bertelli Mottaet al., submitted), we only considered the abundances of the gi-ant stars. The GES solar abundances for the s-process elementsare in very good agreement with the reference solar abundancesfrom Grevesse et al. (2007). The average abundances for thethree member giant stars in M67 (Teff∼4800-4900 and log g∼3-3.4) from the idr5 recommended values are given together withits standard deviation of the mean and the typical error on eachmeasurement (in parenthesis) (see third column of Table1). Inthe following, we normalise our abundances to the solar abun-dances (first column), with the exception of La for which thereis no solar abundance available and thus we make use of the M67abundances.

4. The stellar samples

4.1. The Milky Way field samples

The sample discussed here includes field stars observedwith UVES 580 that belong to the stars in the MilkyWay sample, and in particular to the solar neighbourhoodsample (GES_TYPE = ‘GE_MW’) and the inner disc sample(GES_TYPE = ‘GE_MW_BL’). The selection functions of thesestars are described in Stonkute et al. (2016). For these stars, wecomputed spectroscopic distances as in Magrini et al. (2017), byprojecting the stellar atmospheric parameters and (J-K) 2MASScolours on a set of isochrones (Bressan et al. 2012) spanningthe age range between 0.1 and 13.9 Gyr (with a uniform stepof 0.1 Gyr) and a metallicity range between −2.3 and +0.2 dex(with a uniform step of 0.1 dex). The projection takes into ac-count the uncertainties on the atmospheric parameters, the like-lihood of a star to belong to a given evolutionary phase, and theline-of-sight extinction (iteratively corrected by the distance) ofSchlegel et al. (1998). The projection therefore not only returnsthe absolute magnitude in various bands (and hence, eventually,the line-of-sight distance), but also a crude estimation of the stel-lar ages. The details of the method are described in Kordopatiset al. (2011), with the updates of Recio-Blanco et al. (2014) andKordopatis et al. (2015a). Typical errors ranges from 20% up tomore than 100%, with the bulk of the stars having errors of theorder of 30-50%.

We have divided the field stars into thin and thick disc stars,using their [α/Fe] abundance ratio 1 to discriminate the two pop-ulations, following the approach of Adibekyan et al. (2011). Wehave used the multi slope curve of Adibekyan et al. (2011) todefine thin and thick disc stars on the basis of their [α/Fe] and[Fe/H]. In the following analysis, we have included stars withTeff < 6120 K (to avoid NLTE effects in some neutron-capture el-ement abundances as discussed in Bensby et al. 2014)) and withSNR>70, and applied the error cuts described in the previoussection (see Sec.2) to exclude poor quality results. The adoptedseparation between thin and thick disc stars is also presented inthe first panels of Figures 3 and 4. In the following discussion,we include only thin and thick disc stars located in the Solarneighbourhood (6.5 kpc<RGC <9.5 kpc) to avoid confusion dueto the radial dependence of the abundance ratios.

1 [α/Fe] is computed by averaging [Mg/Fe], [Ca/Fe], [Si/Fe] and ex-cluding [O/Fe], which is affected by large errors in dwarf stars



Table 1. idr5 solar and M67 abundances.

Element Sun (idr5) Sun (Grevesse et al. 2007) M67 giants (idr5)Y II 2.19±0.12 2.21±0.02 2.14±0.01(0.09)Zr II 2.53±0.13 2.58±0.02 2.54±0.03(0.05)Ba II 2.17±0.06 2.17±0.07 2.07±0.03(0.07)La II - 1.13±0.05 1.00±0.02(0.12)Ce II 1.70±0.11 1.70±0.10 1.71±0.01(0.01)Eu II - 0.52±0.06 0.42±0.01(0.04)

4.2. The cluster sample

We consider the sample of open clusters with ages>0.1 Gyrwhose parameters and abundances have been delivered in idr5.We do not include younger clusters because their stars are usu-ally more difficult to analyse and they can be affected by chro-mospheric effects.

For each cluster we have extracted member stars using theinformation on their radial velocities. We have considered asmember stars those within 1.5-σ from the cluster systemic ve-locity (computed using the UVES spectra) and we have excludedoutliers in metallicity |[Fe/H]star− <[Fe/H]> | >0.1 dex. Basedon stars assigned as members, we have computed the medianelemental abundances, expressed in the form [X/H]=log(X/H)-log(X/H)�, which are presented in Table 2. The uncertainties re-ported on each abundance are the standard deviation of clustermember abundances. We consider in the following analysis onlyabundance ratios with dispersion lower than 0.3 dex.

In Table 2 we summarise the basic properties of the sam-ple open clusters: coordinates, ages, Galactocentric distances,heights above the plane, mean radial velocities of cluster mem-bers, median metallicity and the number of cluster member starsused to compute the metallicity. There is a general agreementwith the previous official release (see, e.g. Magrini et al. 2017),but for some clusters the median metallicity might have changedslightly. For clusters in common with Magrini et al. (2017), weadopt the same ages and distances, for the clusters which are newin idr5 we report ages and distances from the recent literature.

In Tables 3 and 4, we present the median abundances[X/H] and abundance ratios, [X/Fe]=[log(X/H)-log(X/H)�)-(log(FeI/H)-log(FeI/H)�)]. The latter may slightly differ fromthe simple subtraction of the median [X/H] and [Fe/H], becausethe mean metallicity [Fe/H] of each cluster is computed fromthe global metallicity (FEH column of idr5), while we havecomputed [X/Fe] using the iron abundance derived from neu-tral iron lines (FE1 column). In principle, for elements withsingly ionised atoms it would be more appropriate to use FeIIabundances to obtain their abundance ratios, [X/Fe]. We havechecked the consistency between log(FeI/H) and log(FeII/H) inour sample, finding an excellent agreement, with a small system-atic offset log(FeII/H)-log(FeI/H)∼0.05 dex which might sys-tematically lower our results by this small amount. However,since the FeII abundances are affected, on average, by larger er-rors due to the smaller number of FeII lines than of FeI lines, weadopted the FeI abundances to compute the [X/Fe] abundanceratios. The choice of FeI to compute [X/Fe] does not affect thetrends or relative comparisons within the Gaia-ESO samples.

5. The abundances of the s-process elements

Following the literature and, in particular, Overbeek et al. (2016)we divide the neutron capture elements into s-elements, thosedominated by s-process if at least 70% of their abundance in theSun is produced by the s-process, and, viceversa, in r-elements,

those dominated by the r-process for more than 70% in the Sun.Elements whose nucleosynthesis is not dominated by either s- orr-process are defined as mixed elements. We follow the review ofSneden et al. (2008) for what concerns the isotopic composition,the mixture of s- and r-processes in the Sun and their relative per-centages, but we compare also with the recent redeterminationof the solar composition by Bisterzo et al. (2014). The isotopiccomposition and (within parenthesis) the total percentage in theSun from Sneden et al. (2008) are reported, for each element, inTable 5. In the last column, we report also the total percentageof each element in the Sun from Bisterzo et al. (2014).

5.1. The effect of the evolutionary stage: comparingabundances in dwarf and giant stars among clustermember stars

In a purely differential analysis with respect to the Sun of stars inopen clusters, D’Orazi et al. (2009) showed that [Ba/Fe] ratiosare systematically higher in giant stars than in dwarf stars in asample of clusters where both kind of stars have been analysed.This effect is related to the use of a dwarf star, the Sun, as areference in the analysis of giant stars.

We have analysed the abundance ratios [X/Fe] as a func-tion of the surface gravity in member stars of clusters in whichboth dwarf and giant stars were observed. The results of [Y/Fe],[Zr/Fe], [Ba/Fe] and [Ce/Fe] for NGC6633, NGC2420, and M67are shown in Figure 1. The behaviour of lanthanum is not shownsince idr5 contains La abundances for almost exclusively giantstars. The aim of Figure 1 is to highlight possible differences inthe abundances of the s-process elements due the different stel-lar parameters. In these plots we have normalized the abundanceratios of the member stars of each clusters to its median abun-dance and we have excluded stars with abundance uncertaintieslarger than 0.2 dex. From this figure, we do not observe, withinthe uncertainties, any systematic trends of the abundance ratio[Y/Fe], [Zr/Fe], [Ce/Fe] with the surface gravity for giant anddwarf member stars of the same cluster. However, for [Ba/Fe]we notice that the stars with very low gravity in NGC2420 haveBa abundance 0.2 dex higher than the other member stars. Inthe range of logg from 2.5 to 4 there is no trend of [Ba/Fe] ver-sus logg and thus no differences between dwarf and giant stars.To probe how much the overestimation of Ba abundance in lowgravity stars might be important, we have selected a sample ofMilky Way field stars with -0.1<[Fe/H]<0.1. In Figure 2 we plottheir [Ba/Fe] versus logg, which indicates that there are no sys-tematic differences between giant and dwarf stars. In addition,In our sample the number of stars with very low gravity (starswith logg<2.5 are ∼3% of the sample of Milky Way field stars)is very small and their eventual over-abundances of Ba shouldnot influence our results.

As a side result, it is worth noticing the presence of a pecu-liar star in NGC2420: one of the member stars of NGC2420 withCNAME=07382696+2133313, has systematically higher abun-



Table 2. Open Cluster parameters

Id R.A. Dec. Age RGC Z rv [Fe/H] n. m. Ref. Age & Distance(J2000.0) (Gyr) (kpc) (pc) (km s−1) (dex)

NGC6067 16:13:11 -54:13:06 0.10±0.05 6.81±0.12 -55±17 -39.0±0.8 +0.20±0.08 12 Alonso-Santiago et al. (2017)NGC2516 07:58:04 -60:45:12 0.11±0.01 7.98±0.01 -97±4 +23.6±1.0 -0.06±0.05 13 Randich et al. (2017)NGC6259 17:00:45 -44:39:18 0.21±0.03 7.03±0.01 -27±13 +33.3±0.9 +0.21±0.04 11 Mermilliod et al. (2001); Dias et al. (2002)NGC3532 11:05:39 -58:45:12 0.30±0.10 7.85±0.01 +11±4 +4.9±0.9 -0.06±0.14 3 Clem et al. (2011)NGC6705 18:51:05 -06:16:12 0.30±0.05 6.33±0.16 -95±10 +34.8±0.7 +0.16±0.04 16 Cantat-Gaudin et al. (2014)NGC6633 18:27:15 +06:30:30 0.52±0.10 7.71±0.01 +52±2 -28.9±0.9 -0.01±0.11 10 Randich et al. (2017)NGC4815 12:57:59 -64:57:36 0.57±0.07 6.94±0.04 -95±6 -29.7±0.5 +0.11±0.01 3 Friel et al. (2014)Tr23 16:00:50 -53:31:23 0.80±0.10 6.25±0.15 -18±2 -61.3±0.9 +0.21±0.04 11 Jacobson et al. (2016)Mel71 07:37:30 -12:04:00 0.83±0.18 10.50±0.10 210±20 +50.8±1.3 -0.09±0.03 5 Salaris et al. (2004)Be81 19:01:36 -00:31:00 0.86±0.10 5.49±0.10 -126±7 +48.2±0.5 +0.22±0.07 13 Magrini et al. (2015)NGC6802 19:30:35 +20:15:42 1.00±0.10 6.96±0.07 +36±3 +11.9±0.9 +0.10±0.02 8 Jacobson et al. (2016)Rup134 17:52:43 -29:33:00 1.00±0.20 4.60±0.10 -100±10 -40.9±0.6 +0.26±0.06 17 Carraro et al. (2006)NGC6005 15:55:48 -57:26:12 1.20±0.30 5.97±0.34 -140±30 -23.5±1.0 +0.19±0.02 11 Piatti et al. (1998)Pis18 13:36:55 -62:05:36 1.20±0.40 6.85±0.17 +12±2 -27.5±0.6 +0.22±0.04 6 Piatti et al. (1998)Tr20 12:39:32 -60:37:36 1.50±0.15 6.86±0.01 +136±4 -40.1±1.0 +0.15±0.07 37 Donati et al. (2014)Be44 19:17:12 +19:33:00 1.60±0.30 6.91±0.12 +130±20 -8.8±0.5 +0.27±0.06 7 Jacobson et al. (2016)NGC2420 11:05:39 -58:45:12 2.20±0.30 10.76±0.20 +765±50 +4.9±0.9 -0.13±0.04 22 Salaris et al. (2004); Sharma et al. (2006)Be31 06:57:36 +08:16:00 2.50±0.30 15.16±0.40 +340±30 +57.5±0.9 -0.27±0.06 9 Cignoni et al. (2011)Be25 06:41:16 -16:29:12 4.00±0.50 17.60±1.00 -1900±200 +136.0±0.8 -0.25±0.05 6 Carraro et al. (2005)NGC2243 06:29:34 -31:17:00 4.00±1.20 10.40±0.20 -1200±100 +60.2±0.5 -0.38±0.04 16 Bragaglia & Tosi (2006)M67 08:51:18 +11:48:00 4.30±0.50 9.05±0.20 +405±40 +34.7±0.9 -0.01±0.04 19 Salaris et al. (2004)Be36 07:16:06 -13:06:00 7.00±0.50 11.3±0.20 -40±10 +62.3±1.6 -0.16±0.10 7 Donati et al. (2012)

Table 3. Abundance in open clusters, expressed in the form [X/H].

Id [YII/H] [ZrII/H] [CeII/H] [BaII/H] [LaII/H] [EuII/H]NGC6067 0.19±0.05 0.36±0.07 0.13±0.07 0.46±0.09 0.20±0.06 0.19±0.06NGC2516 0.04±0.05 0.52±0.06 0.28±0.10 0.17±0.08 - -NGC6259 0.14±0.03 0.28±0.08 0.09±0.12 0.08±0.05 0.13±0.11 0.20±0.04NGC3532 -0.09±0.24 - - 0.02±0.24 - -NGC6705 0.05±0.03 0.22±0.06 0.05±0.09 0.20±0.11 0.06±0.06 0.15±0.05NGC6633 0.01±0.04 0.15±0.01 0.09±0.02 0.12±0.11 0.06±0.03 -0.03±0.13NGC4815 0.08±0.07 0.23±0.10 0.06±0.08 0.24±0.10 0.09±0.07 0.05±0.08Tr23 0.04±0.04 0.22±0.04 0.00±0.10 0.15±0.13 -0.01±0.04 0.16±0.08Mel71 -0.09±0.01 0.07±0.02 0.04±0.04 0.17±0.05 -0.01±0.03 -0.09±0.05Be81 0.23±0.05 0.30±0.11 0.19±0.12 0.01±0.08 0.24±0.09 0.21±0.08NGC6802 0.16±0.02 0.34±0.06 0.17±0.07 0.16±0.05 0.13±0.06 0.11±0.02Rup134 0.16±0.02 0.32±0.08 0.15±0.08 0.04±0.05 0.17±0.05 0.22±0.04NGC6005 0.12±0.01 0.24±0.03 0.13±0.01 0.08±0.06 0.03±0.04 0.12±0.04Pis18 0.06±0.10 0.25±0.08 0.05±0.03 0.14±0.08 0.09±0.10 0.16±0.02Tr20 0.13±0.03 0.25±0.05 0.13±0.09 0.12±0.07 0.10±0.05 0.13±0.07Be44 0.34±0.05 - - 0.05±0.18 0.23±0.07 0.18±0.07NGC2420 -0.14±0.04 0.01±0.06 0.04±0.05 0.04±0.09 -0.01±0.04 -0.06±0.08Be31 -0.32±0.05 -0.11±0.08 -0.09±0.09 -0.32±0.02 -0.14±0.07 -0.16±0.04Be25 -0.24±0.13 0.17±0.21 0.01±0.22 -0.28±0.23 0.03±0.11 0.00±0.09NGC2243 -0.41±0.08 -0.21±0.06 -0.34±0.07 - -0.34±0.08 -0.31±0.10M67 -0.07±0.02 0.08±0.08 0.01±0.03 0.00±0.09 0.16±0.19 -0.10±0.09Be36 -0.18±0.04 -0.10±0.04 -0.11±0.05 - -0.10±0.20 -0.01±0.07

dances of Y, Zr, and Ba. Its radial velocity is perfectly consistentwith those of the other cluster members. However, in the HRdiagram it is located slightly outside the main locus of the RedClump stars and also the abundance of some other elements aresomewhat discrepant with respect to the average abundance ofthe cluster (for instance [O/H], [C/H], [Si/H], [Sc/H], [Ti/H] areabout -0.3 dex, -0.1 dex, -0.1 dex, -0.15 dex, -0.15 dex respec-tively, lower than the average cluster values). Since its [Fe/H]and radial velocity are consistent with those of the cluster, wecan make several hypothesis on the nature of this star: i) it is nota cluster member, but a field star with exactly the same metallic-ity and radial velocity of the cluster; ii) it is a remnant of binarysystem in which one of two companions was merged thus mod-ifying the surface composition of the other companion; iii) thedifferent composition might be due to planet engulfment (see,

e.g, Spina et al. 2015). However, the study of this peculiar staris beyond the scope of the present paper and it deserves furtherinvestigations.

5.2. The origin of the s-process elements

Most of the neutron capture elements are produced in a mix-ture of astrophysical environments, involving both the r- and s-processes. There are a few cases in which one of the two pro-cesses dominates above the other, being then the major processresponsible for the element production, but in general both pro-cesses contribute, with one or the other dominating at differentepochs and thus in different metallicity ranges.



Table 4. Abundance ratios in open clusters, expressed in the form [X/Fe].

Id [YII/Fe] [ZrII/Fe] [CeII/Fe] [BaII/Fe] [LaII/Fe] [EuII/Fe]NGC6067 0.14±0.09 0.20±0.10 0.07±0.10 0.39±0.11 0.13±0.09 0.16±0.10NGC2516 0.11±0.06 0.58±0.07 0.38±0.10 0.20±0.08 - -NGC6259 -0.01±0.05 0.13±0.08 -0.06±0.12 -0.04±0.06 -0.01±0.11 -0.05±0.06NGC3532 0.06±0.13 - - 0.17±0.11 - -NGC6705 0.03±0.06 0.15±0.08 -0.01±0.10 0.15±0.12 0.03±0.08 0.01±0.08NGC6633 0.11±0.10 0.24±0.09 0.18±0.09 0.24±0.14 0.15±0.10 -0.05±0.16NGC4815 0.11±0.12 0.29±0.14 0.13±0.12 0.31±0.13 0.12±0.12 -0.05±0.13Tr23 0.01±0.05 0.22±0.05 -0.01±0.11 0.05±0.14 -0.03±0.05 0.00±0.09Mel71 0.08±0.02 0.23±0.02 0.21±0.04 0.34±0.05 0.16±0.03 -0.05±0.06Be81 0.10±0.08 0.19±0.12 0.05±0.14 -0.10±0.10 0.11±0.10 0.00±0.10NGC6802 0.15±0.06 0.36±0.08 0.18±0.09 0.16±0.07 0.15±0.08 -0.01±0.06Rup134 -0.04±0.06 0.14±0.09 -0.02±0.10 -0.15±0.08 -0.06±0.08 -0.08±0.07NGC6005 -0.01±0.01 0.12±0.03 0.03±0.01 -0.03±0.05 -0.07±0.04 -0.14±0.04Pis18 0.06±0.10 0.20±0.07 0.04±0.03 0.08±0.08 0.06±0.10 0.00±0.02Tr20 0.08±0.07 0.20±0.08 0.10±0.11 0.08±0.09 0.04±0.08 -0.03±0.09Be44 0.13±0.07 - 0.06±0.10 - 0.05±0.09 -0.12±0.09NGC2420 0.07±0.05 0.22±0.07 0.21±0.06 0.23±0.09 0.18±0.05 0.00±0.08Be31 0.05±0.05 0.28±0.09 0.31±0.10 0.06±0.04 0.23±0.08 0.09±0.06Be25 0.11±0.22 0.30±0.27 0.34±0.28 0.05±0.29 0.21±0.20 0.19±0.19NGC2243 0.07±0.09 0.28±0.08 0.16±0.09 - 0.19±0.10 0.12±0.12M67 -0.01±0.04 0.14±0.08 0.05±0.05 0.08±0.10 0.24±0.19 -0.19±0.10Be36 0.05±0.20 0.18±0.20 0.10±0.20 - 0.11±0.27 0.13±0.20

Table 5. Solar percentages and isotopic composition for the s-process elements from Sneden et al. (2008) and Bisterzo et al. (2014).

Element Isotopes Sneden (2008) Bisterzo et al. (2014)Y 89Y 72% 72%Zr 90Zr, 91Zr, 92Zr, 94Zr, 96Zr 52%,11%, 15%, 17%, 3% (81%) 66%La 139La 75% 76%Ce 140Ce, 142Ce 90%, 10% (81%) 84%Ba 134Ba, 135Ba, 136Ba, 137Ba, 138Ba 3%, 7%, 9%, 12%, 69% (85%) 85%

In Figures 3 and 4, we show the abundance ratios [X/Fe]of the s-process elements ordered by increasing atomic numberplotted with respect to [Fe/H]. Figure 3 presents the thin andthick individual stars abundances and the open cluster medianabundances: this figure allows us to estimate the differences be-tween field stars and open clusters. We note that the age and dis-tance distributions of the cluster and field star samples plottedhere differ in important ways. The majority of field stars haveages between 3 and 7 Gyr, with very few stars younger than 2Gyr. By contrast, all but six of the open clusters are youngerthan 2 Gyr. In addition, the fields stars have been limited to 1.5kpc around the solar radius, from 6.5 kpc<RGC <9.5 kpc, whilethe open clusters range from 4.5 to 18 kpc in Galactocentric ra-dius. Consequently, any age and distance dependent trends inabundances will affect the cluster and field star distributions inthese diagrams differently. To help understand the impact of thedistance differences, we have color-coded the open clusters bytheir Galactocentric distance, separating the clusters into radialranges of RGC <6.5 kpc (blue), 6.5 kpc<RGC <9.5 kpc (green),and RGC >9.5 kpc (violet).

Figure 4 shows the thin and thick abundances, averaged inbins 0.1 dex wide: this figure aims at comparing the differ-ent Galactic populations and to seek differences between thickand thin disc stars. The binned results are shown in the Fig-ures only if there is more than one star in the considered in-terval. The errors on the binned values are 1-sigma standarddeviation of the average. Typical errors on the individual abun-dances are: 0.15 dex for log(Y II/H), 0.15 dex for log(Zr II/H),0.08 dex for log(Ba II/H), 0.09 dex for log(La II/H), 0.14 dex forlog(Ce II/H), and 0.16 dex for log(Eu II/H).

In the first panel of both Figures 3 and 4, we present the[α/Fe] vs. [Fe/H] distribution of field stars, with the chemicaldividing line between thin and thick disc stars. To separate thinand thick stars we have used the relationship of Adibekyan et al.(2011). As expected from their young ages, most open clustersbelong to the thin disc populations. The separation between thethin and thick disc stars is clear and in agreement with previousresults (e.g. Recio-Blanco et al. 2014).

In the second panels, [Y/Fe] is almost flat across the wholemetallicity range, with a relatively small dispersion and a slightdecrease towards the lower metallicity. The cluster abundancesoccupy the upper envelope of the region defined by the abun-dances of field stars, while the distributions of thin and thickdisc stars are coincident (second panel of Figure 3). This differ-ence between cluster and field stars is seen here for the first timewith a statistically significant sample of cluster members and offield stars. In the next sections, we show that this is likely anevolutionary effect due to the younger ages of cluster stars withrespect to the field population.

In the third panels, [Zr/Fe] increases both in thin and thickdisc towards lower metallicity, indicating a non-negligible con-tribution from massive stars, which, given its higher s-processpercentage with respect to Y, is a bit unexpected. The recent pa-per by Delgado Mena et al. (2017) also indicates that Zr shows aclear increasing trend for lower metallicities. The trend is notnew, and it was observed in previous works (e.g., GonzálezHernández et al. 2010; Mishenina et al. 2013; Battistini &Bensby 2016; Zhao et al. 2016). The behaviour of Zr can be re-lated to a reassessment of its s-process contribution. The work ofBisterzo et al. (2014) indicates that only ∼66% of Zr is producedby the s-process in the Sun. Other processes may contribute to



Fig. 1. Abundance ratios [X/Fe] normalised to the median abundance of the cluster versus logg in member stars of the open clusters NGC2420(yellow), NGC6633 (red), and M67 (green). Stars with errors in the abundances larger than 0.2 dex are not included in the plot.

the production of Zr, as discussed by Travaglio et al. (2004) whointroduced the LEPP process associated with massive stars toexplain the abundance patterns of the neutron capture elements,especially at low metallicity. As in the case of [Y/Fe], [Zr/Fe]in open clusters is slightly higher with respect to thin and thickstars, which appear statistically indistinguishable, although bothclusters and field stars have [Zr/Fe] elevated above the solar ra-tio.

With a percentage of 85%, Ba (fourth panels of Figures 3 and4) is dominated by the s-process in the metallicity range from[Fe/H]≥-1.0 to super-solar metallicity. Only at lower metallic-ities, the primary r-process contribution, relatively small at thetime of formation of the Sun, plays an important role (Travaglioet al. 1999). In the field stars, the abundance ratio [Ba/Fe] is closeto solar for metallicities between [Fe/H]=-1.0 and 0, with a peakof [Ba/Fe]∼0.1 dex in the range -0.4<[Fe/H]<-0.6. In the super-solar metallicity range, we distinguish between [Ba/Fe] in openclusters and field stars; [Ba/Fe]≥0 dex for clusters located in theSolar neighbourhood and it is negative for the innermost, moremetal rich clusters and for the thin and thick disc stars. The thickdisc abundance ratios are systematically lower than the thin discratios, with the exception of stars with [Fe/H]>0 dex for whichthin and thick disc stars have coincident [Ba/Fe] distributions.

Among the second-peak elements, La (fifth panels of Fig-ures 3 and 4) has a slight increase of [La/Fe] towards low metal-licity. The open clusters are located in the upper envelope of thedistribution of field star abundance ratios. Thin and thick discstars show similar patterns, with thin disc stars slightly overabun-dant in [La/Fe].

Finally, the behaviour of [Ce/Fe] (sixth panels of Figures 3and 4) is almost flat in the [Fe/H] range considered. At vari-ance with [Ba/Fe], [Ce/Fe] is constant also at super solar metal-licities. Thin and thick stars share similar patterns.For both ele-ments, Ce and La, the more metal-poor open clusters have higher[La/Fe] and [Ce/Fe] than the more metal-rich ones. However,these metal-poor clusters are located in the outer disc and theirabundances may be more a reflection of the radial gradient ofthese elements in the Galactic disc.

5.3. Estimating the r-process contribution

The abundance ratio [X/Eu] can be used as a direct way to judgethe relative importance of the r-process channels during the evo-lution of the Galaxy in the metallicity range of our sample stars,as has been done in previous works (e.g., Mashonkina & Gehren2001; Battistini & Bensby 2016). In Figures 5 and 6 we presentthe [X/Eu] abundance ratios versus [Fe/H]. In the plots the blue



Fig. 2. Abundance ratios [Ba/Fe] versus logg in Milky Way field starswith -0.1<[Fe/H]<0.1.

dashed lines represent the s-process solar values, while the reddot-dashed line indicates the pure r-process contribution of bothelements, derived using the percentages of Bisterzo et al. (2014)and the solar abundances of Grevesse et al. (2007). The two fig-ures highlight different aspects: in the former we have a directcomparison between field stars and open clusters, and in the lat-ter the comparison between thin and thick disc populations.

In the first panels of both figures, we present the abundanceratio [Eu/Fe] versus [Fe/H]. Europium is an r-process elementand it shows a clear increase towards the lower metallicity. Al-though the first hypothesis about the mechanisms responsible ofr-process nucleosynthesis date back many years ago (Burbidgeet al. 1957; Seeger et al. 1965), there is still an open debate aboutthe dominant production site of the r-process elements (see e.g.Thielemann et al. 2011). The r-process nucleosynthesis must beassociated with both the evolution of massive stars (M >10 M�)and with neutron star mergers (Matteucci et al. 2014). Very re-cently, observations of the optical counterpart of a gravitationalwave source have identified the presence of r-process elementsin the spectrum of the merger of two neutron stars, clearly point-ing to the production of these elements in neutron star mergerevents (Côté et al. 2017).

In the second panels, for the first-peak element Y the sampleshows an increasing contribution from s-process elements abovea pure r-process level, but there are some stars in the range -1.0<[Fe/H]<-0.5 that appear to have a pure r-process composi-tion. These stars belong to both the thin and thick discs. Over thefull metallicity range, the abundance ratios of the thick and thindiscs are offset from each other by about 0.1 dex in [Y/Eu].

In the third panels, we present the abundance ratio [Zr/Eu]versus [Fe/H], which is almost constant, quite far from the purer-process contribution, and slightly above the solar value. Al-most all the open clusters have positive [Zr/Eu]. For this el-ement, the thin and thick disc populations are separated: theabundance ratios of thick disc stars are lower, although they fol-low a similar pattern to the thin disc stars (see third panel ofFigure 6). In the fourth panels, barium shows a mild increaseof the r-process component towards low metallicity. In the thindisc [Ba/Eu] is close to solar values from [Fe/H]∼+0.4 down to[Fe/H]∼-0.5 dex, then it decreases. For the thick disc, the decline

starts at higher metallicity, [Fe/H]∼-0.2 dex, and again, over theentire metallicity range, the thick disk stars have lower abun-dance ratio. [Ba/Eu] in open clusters is positive in the range -0.2<[Fe/H]<+0.2 dex, while it is close to zero for the clustersmore metal poor and more metal rich.

In the fifth panels, [La/Eu] is quite flat over the whole metal-licity range. The open clusters are all slightly over solar in[La/Eu], and thick disk abundance ratios are slightly lower thanin thin disk stars.

In the sixth panels, [Ce/Eu] is also quite flat over the wholemetallicity range and it tends to increase in both thin and thickdisc stars towards high [Fe/H]. The few stars at low metallicity[Fe/H]≤-0.5 dex might point to a decrease of [Ce/Eu], indicatingthe beginning of the r-process contribution in the thick disc.

We can thus conclude that in the considered metallicityrange, the production of the five elements is clearly dominatedby the s-process. At any metallicity, the thick disk stars showvery slightly lower [s/Eu] ratios than the thin disk, consistentwith their slightly higher [Eu/Fe] values.

6. The time evolution of the abundances of thes-process elements

One of the most challenging issues in the field of chemical evo-lution is the nature of the time-evolution of the s-process abun-dances. The classical chemical and chemo-dynamical models(Pagel & Tautvaisiene 1997; Travaglio et al. 1999, 2004; Rai-teri et al. 1999), based on the accepted scenario for s-processingat that time, predicted a plateau or even a decrease in the abun-dances [X/Fe] versus age after the solar formation. A late in-crease was not expected. However, recently, after the originalwork of D’Orazi et al. (2009), a number of works have claimedthe presence of a continuous increase of the first- and second-peak element abundances, sometimes with a further growth start-ing 5-6 Gyr ago (Maiorca et al. 2011; da Silva et al. 2012; Nissen2016; Spina et al. 2016; Nissen et al. 2017; Silva Aguirre et al.2017; Feltzing et al. 2017).

In the present work, we combine the open cluster sample,which basically maps the recent epochs in the evolution of theGalactic disc, with the field stars, divided into thin and thickdisc on the basis of their [α/Fe], as explained in Section 4.1.The ages of the field stars, computed with the isochrone projec-tion methods, are much less accurate than the ages of open clus-ters, and they have only a statistical meaning. However, keep-ing in mind the limit of measuring stellar ages, the Gaia-ESOidr5 sample gives us the opportunity, for the first time, to com-bine the largest sample of s-process element abundances mea-sured in a completely homogeneous way in very young and veryold stars. No biases due to, for instance, the line list selection,the method of deriving abundances, or adopted reference abun-dances are present in the Gaia-ESO idr5.

In Figure 7, 8, and 9 we show the abundance ratios [X/Fe]versus age of both field stars and open clusters. Recall that thesample of thin and thick disk stars is already restricted to the So-lar neighbourhood, i.e. 6.5 kpc<RGC <9.5 kpc. The open clustersspan a much larger range of distances, and we have distinguishedthem by radial bins according to their Galactocentric location,denoting inner disk, solar neighbourhood, and outer disk clustersseparately in the figures. In Figure 7 we compare the field starswith the open clusters, while in Figure 8 we compare the thin andthick disc [X/Fe] versus stellar ages (binning in age bins 1 Gyrwide), and finally in Figure 9 we limit our analysis to the Solarneighbourhood thin disc populations, and open clusters belong-ing to the same radial range.



Fig. 3. Abundance ratios [X/Fe] versus [Fe/H] in member stars of clusters (large filled circles– in blue for clusters with RGC <6.5 kpc, in green forclusters with 6.5 kpc<RGC <9.5 kpc, and in violet for clusters with RGC >9.5 kpc), in thin disc stars (small grey squares) and in thick disc stars(small black triangles). The blue dashed lines represent the solar values. The red dash-dotted line is the relationship of Adibekyan et al. (2011)which separates thin and thick disc stars.

Fig. 4. Abundance ratios [X/Fe] versus [Fe/H] in thin disc stars (cyan squares, binned results, bin width 0.1 dex) and in thick disc stars (orangetriangles, binned results, bin width 0.1 dex). The blue dashed lines represent the solar values. The red dash-dotted line is the relationship ofAdibekyan et al. (2011) which separates thin and thick disc stars.



Fig. 5. [Eu/Fe] and abundance ratios [X/Eu] versus [Fe/H]. Symbol and colour codes as in Fig. 3. The blue dashed lines represent the solarvalues, while the red dot-dashed lines indicate the abundance ratio below which the pure r-contribution dominates, derived using the percentagesof Bisterzo et al. (2014) and the solar abundances of Grevesse et al. (2007).

Fig. 6. [Eu/Fe] and abundance ratios [X/Eu] versus [Fe/H]. Symbol and colour codes as in Fig. 4. The lines are coded as in Figure 5.



Figure 7 aims at highlighting the behaviour of clusters lo-cated at different Galactocentric distance with respect to the fieldstars. The Solar neighbourhood and inner disc clusters populatethe youngest age regions, while the outer disc clusters span alarge range of ages. For all s-process elements, we notice a dif-ference between the youngest clusters of the Solar neighbour-hood and the inner disc clusters. The former usually have higher[X/Fe] than the latter. The outer disc open clusters typically have[X/Fe] similar to the Solar neighbourhood clusters. Dividing thecluster sample by Galactocentric distance makes clear that thereis a strong dependence of the heavy s-element abundances on thelocation of the clusters. The behaviour of [X/Fe] with Galacto-centric radius seems to be complex and likely related to differentstar formation histories and to metallicity dependency of the stel-lar yields.

Figure 8 shows [X/Fe] as a function of stellar ages for thebinned thin and thick disc populations. In the radial range shownhere, i.e. 6.5 kpc <RGC <9.5 kpc, there are not strong differencesbetween the behaviours of the thin and thick disc populations.All elements seem to indicate a slight increase in [X/Fe] start-ing at ages <8 Gyr. These trends are most clear in the thin disk,where the sample sizes per bin are larger and mean values areperhaps better determined than for the thick disc.

Figure 9 presents [X/Fe] versus stellar age for the thindisc and cluster populations in the Solar neighbourhood withweighted linear fits to the combined cluster and thin disc samplesfor ages lower than 8 Gyr, i.e. corresponding to epoch in whichwe expect the contribution of the s-process starts. In Table 6 wepresent these weighted linear fits, reporting the slopes, the inter-cepts, and correlation coefficients for fits by element. Restrictingour sample to the Solar neighborhood allows us to better appre-ciate the increasing trends with decreasing stellar ages. The firstpeak elements, [Y/Fe] and [Zr/Fe], have increases of 0.15 to 0.25dex over the past 6-7 Gyr. These slopes agree with those of Spinaet al. (2018), who find slopes of −0.029 and −0.027 dex Gyr−1for Y and Zr, respectively. The elements of the second peak alsoshow a slight increase with age, more pronounced for [Ba/Fe]and [Ce/Fe], and flatter for [La/Fe] (see Table 6). Slopes for Ba,at −0.03 dex Gyr−1, agree with those found by da Silva et al.(2012); Spina et al. (2018). In the case of La, which is measuredonly in giant stars, we have a much smaller sample of stars andthus the measurement of the trend is less well constrained (seeits relative error on the slope and its correlation coefficient inTable 6).

The growth in the s-process abundances in the youngestGalactic stellar populations has been reproduced by, for instance,the chemical evolution model of Maiorca et al. (2012), whichwas one of the first models that made an attempt to explainit. Their approach was to relate the s-process enhancement toa strong contribution to the production of these elements fromlow-mass stars, which start to contribute later in the lifetime ofthe Galaxy. In particular, they maintain that the observed en-hancements might be produced during the nucleosynthesis pro-cesses in the asymptotic giant branch (AGB) phases of low-massstars (M <1.5 M�) under the hypothesis that they release neu-trons from the 13C(α,n)16O reaction to create larger reservoirs ofneutrons than in more massive AGB stars (M >1.5 M�). Laterstudies have given a physical explanation for the necessity ofhaving larger reservoirs of neutrons, the so-called 13C pocket,in low-mass stars. The extended 13C pocket requires the exis-tence of very efficient mixing episodes and the transport mecha-nisms most commonly adopted such as overshoot (Herwig 2000;Cristallo et al. 2009), gravity waves (Denissenkov et al. 2013;Battino et al. 2016), rotationally driven shear or thermohaline

Table 6. Coefficients and Pearson indexes of the weighted linear fits for[X/Fe] versus stellar age (age<8 Gyr).

Element Slope (dex Gyr−1) Intercept (dex) RYttrium [Y/Fe] -0.023±0.009 0.13±0.03 -0.6Zirconium [Zr/Fe] -0.038±0.013 0.26±0.03 -0.5Barium [Ba/Fe] -0.027±0.007 0.09±0.02 -0.8Lanthanum [La/Fe] -0.005±0.015 0.10±0.04 -0.2Cerium [Ce/Fe] -0.016±0.010 0.10±0.03 -0.5

mixing do not suffice (Palmerini et al. 2011), but magnetic buoy-ancy can explain the process (Busso et al. 2007; Nordhaus et al.2008; Trippella et al. 2016; Trippella & La Cognata 2017).

The observed slope of [Ba/Fe] as a function of stellar age issomewhat steeper than those of [La/Fe] and [Ce/Fe], althoughtheir intercepts indicate similar maximum global enrichment (∼0.1 dex) for all elements, in agreement with their similar nu-cleosynthesis. However the maximum abundances of [Ba/Fe]reached in the youngest open clusters is higher than [La/Fe] and[Ce/Fe]. A larger enrichment of Ba compared to La is difficultto reconcile considering similar contributions to both elementsby s- and r-processes. Mishenina et al. (2015) discussed severalpossibilities to explain [Ba/La] in young open clusters, whichcan reach ∼0.15 dex. They proposed an additional contributionfrom an intermediate neutron capture process, the so-called i-process. The i-process is characterised by neutron densities inthe order of 1015 neutrons cm−3 and is triggered by the mixingor ingestion of H in He-burning stellar layers (Cowan & Rose1977; Bertolli et al. 2013). The i-process can potentially explainthis larger ratio and the higher [Ba/La] in the youngest objectsmight suggest that the i-process contribution is becoming morerelevant during the most recent few Gyr.

6.1. The s-process relative indicator [Ba/Y]

In Figure 10 we show [Ba/Y] versus stellar ages. Y and Ba be-long to the first and second peak, respectively, and thus they al-low us to estimate the relative weight of the production of lightand heavy s-process elements with time. We have chosen thesetwo elements because they are measured in many more stars thanZr, La, and Ce. In the plot we compare our results with the sam-ple of solar twin stars of Spina et al. (2018). To do that, we haveselected thin disc stars and open clusters with the same character-istics of the Spina et al. (2018)’s sample: 6.8 kpc<RGC < 8.2 kpcand -0.1<[Fe/H]< +0.1. The two samples are in very good agree-ment, both indicate for [Ba/Y] an increase from about ∼6 Gyrto the youngest ages. The oldest stars show a slight growth of[Ba/Y] together with an increasing dispersion. In this Figure,we have also plotted the results of the model of Maiorca et al.(2012) for the time evolution of [Y/Ba]. The model results are inqualitative agreement, whilst it predicts a lower [Ba/Y] for ageslarger of ∼6 Gyr. As already noticed by Spina et al. (2018), thebehaviour of [Ba/Y] versus stellar age is directly related to de-pendence of the AGB star yields on metallicity. The interplayduring the chemical evolution of the Milky Way between verylow-mass AGB stars, with enhanced reservoir of 13C, and moremassive AGB stars with the usual 13C pocket, might produce theobserved behaviour of [Ba/Y] (see, e.g., Figure 3 of Maiorca etal. (2012)). The observational result can be explained also by theAGB models of Karakas & Lugaro (2016), which predict thatthe [hs/ls] ratio in the winds of a typical AGB star is -0.026 atsolar metallicity and +0.320 at half solar metallicity.



Fig. 7. Abundance ratios [X/Fe] versus Age. Symbol and colour codes as in Fig. 3.

Fig. 8. Abundance ratios [X/Fe] versus Age. Symbol and colour codes as in Fig. 4.



Fig. 9. Abundance ratios [X/Fe] versus Age for the thin disc stars (binned results, bin 0.1 dex wide, cyan squares) and the open clusters locatedin the solar neighborhood (green circles). The magenta dashed lines are the weighted linear fits to the cumulative sample of solar neighbourhoodclusters and thin disc field stars (see Table 6 for the coefficients) for stellar ages <8 Gyr.

In Figure 11, we expand the study of the behaviour of [Ba/Y]versus stellar ages to different radial and metallicity bins. We ba-sically divide the sample in three radial ranges: the solar neigh-bourhood 6.2 kpc<RGC < 8.2 kpc, the inner disc RGC < 6.2 kpcand the outer disc RGC > 8.2 kpc. Each radial range is thendivided in metallicity bins: below solar [Fe/H]< -0.1, solar -0.1<[Fe/H]< +0.1 and super solar [Fe/H]> +0.1. The relativeweight of the s-process indicator [Ba/Y] clearly depends on theGalactocentric position, and thus on the infall time scales andstar formation history, and on the metallicity. The growth of[Ba/Y] with stellar ages is present in the solar neighbourhood,but it is less clear and less constrained in the inner disc andin outer disc. The metallicity dependence is very strong, with[Ba/Y] usually higher in the [Fe/H]< -0.1 range, and decreasingtowards the higher metallicities. All this information is new, andit was not available at the time of previous studies. It providesimportant constraints to the study of the s-process production.

7. Summary and conclusions

Taking advantage of the large and homogeneous sample of stel-lar parameters and elemental abundances of Gaia-ESO idr5, westudy the abundances of five s-process elements in the thin andthe thick discs and in the population of open clusters. We com-pute statistical ages for field stars and we perform membershipanalysis of stars in open clusters, computing the median abun-dances of each cluster. We estimate the effect of evolutionarystages in the s-process abundances by comparing giant and dwarfstars, which are members of the same clusters, and we do not findany remarkable difference. We identify the abundance ratio pat-

Fig. 10. Abundance ratios [Ba/Y] versus Age in cluster (green circles)and thin disc stars (cyan squares, bins of 1 Gyr) in the radial range6.8 kpc<RGC < 8.2 kpc and -0.1<[Fe/H]< +0.1, and in the solar twinstars of Spina et al. (2018) (purple stars). In red the model curve ofMaiorca et al. (2012).

terns versus [Fe/H] of both field and open cluster stars, and theirbehaviour versus stellar ages. Using our large and homogenoussample of open clusters, thin and thick disc stars, spanning anage range larger than 10 Gyr, and selecting a sample of clusters



Fig. 11. Abundance ratios [Ba/Y] versus Age in cluster and thin disc stars (symbols as in Figure 10 ) the three radial ranges: 6.2 kpc<RGC < 8.2 kpc(left panels), RGC < 6.2 kpc (central panels) and RGC > 8.2 kpc (right panels). For each radial bin, we divide the plot in three metallicity ranges:below solar [Fe/H]< -0.1 (panels on the upper row), solar -0.1<[Fe/H]< +0.1 (panels on the central row) and super solar [Fe/H]> +0.1 (panels onthe lower row).

and field stars in the solar ring, we confirm an increase towardsyoung ages of the five analysed s-process abundances. The trendis more clear for [Y/Fe], [Zr/Fe] and [Ba/Fe], while it has a lowercorrelation coefficient for [La/Fe] and [Ce/Fe] perhaps becauseof the limited number of stars in which these elements are mea-sured. Clusters located in the inner and outer disc have differentabundance ratios with respect to those of the solar neighbour-hood.

The global growth of the s-process abundance ratio with timecan be explained by a strong contribution to the production ofthese elements from low-mass stars, which start to contributelater in the lifetime of the Galaxy (see, e.g. Maiorca et al. 2012).However, the differences in the maximum values reached in openclusters for Ba and La point to the necessity of an additionalproduction mechanism for Ba and they might be explained bythe occurrence of the i-process at later epochs in the evolutionof the Galaxy (e.g. Mishenina et al. 2015). The new observa-tions of Gaia-ESO have confirmed with a large statistical samplethe behaviour of s-process elements in the Solar neighbourhood,and now they give us the possibility to investigate the interplayamong the different neutron-capture processes, the metallicitydependence of their stellar yields, and the physical process in-volved in the creation of this important group of elements in dif-ferent parts of our Galaxy.

Acknowledgements. We are grateful to the referee for her/his comments andsuggestions that improved the quality of the paper. These data products havebeen processed by the Cambridge Astronomy Survey Unit (CASU) at the In-stitute of Astronomy, University of Cambridge, and by the FLAMES/UVESreduction team at INAF/Osservatorio Astrofisico di Arcetri. These data havebeen obtained from the Gaia-ESO Survey Data Archive, prepared and hosted

by the Wide Field Astronomy Unit, Institute for Astronomy, University of Ed-inburgh, which is funded by the UK Science and Technology Facilities Coun-cil. This work was partly supported by the European Union FP7 programmethrough ERC grant number 320360 and by the Leverhulme Trust throughgrant RPG-2012-541. We acknowledge the support from INAF and Ministerodell’ Istruzione, dell’ Università’ e della Ricerca (MIUR) in the form of thegrant "Premiale VLT 2012". The results presented here benefit from discus-sions held during the Gaia-ESO workshops and conferences supported by theESF (European Science Foundation) through the GREAT Research NetworkProgramme. F.J.E. acknowledges financial support from ASTERICS project(ID:653477, H2020-EU.1.4.1.1. - Developing new world-class research infras-tructures). S. D. acknowledges support from Comité Mixto ESO-GOBIERNODE CHILE. AB thanks for support from the Millenium Science Initiative,Chilean Ministry of Economy. This project has received funding from the Eu-ropean Union’s Horizon 2020 research and innovation programme under theMarie Sklodowska-Curie grant agreement No 664931. E.D.M., V.A. and S.G.S.acknowledge the support from Fundação para a Ciência e a Tecnologia (FCT,Portugal) through the research grant through national funds and by FEDERthrough COMPETE2020 by grants UID/FIS/04434/2013 & POCI-01-0145-FEDER-007672, PTDC/FIS-AST/1526/2014 & POCI-01-0145-FEDER-016886and PTDC/FIS-AST/7073/2014 & POCI-01-0145-FEDER-016880. E.D-M.,V.A. and S.G.S also acknowledge support from FCT through Investigador FCTcontracts nr. IF/00849/2015/CP1273/CT0003, IF/00650/2015/CP1273/CT0001and IF/00028/2014/CP1215/CT0002. T.B. was supported by the project grant’The New Milky’ from the Knut and Alice Wallenberg foundation. T. M. ac-knowledges support provided by the Spanish Ministry of Economy and Compet-itiveness (MINECO) under grant AYA-2017-88254-P

ReferencesAdibekyan, V. Z., Santos, N. C., Sousa, S. G., & Israelian, G. 2011, A&A, 535,

LL11Alonso-Santiago, J., Negueruela, I., Marco, A., et al. 2017, MNRAS, 469, 1330Asplund, M., Grevesse, N., Sauval, A. J. & Scott, P. 2009, ARA&A, 47, 481



Battino, U., Pignatari, M., Ritter, C., et al. 2016, ApJ, 827, 30Battistini, C., & Bensby, T. 2016, A&A, 586, A49Bensby, T., Feltzing, S., & Oey, M. S. 2014, A&A, 562, A71Bertolli, M. G., Herwig, F., Pignatari, M., & Kawano, T. 2013, arXiv:1310.4578Bisterzo, S., Travaglio, C., Gallino, R., Wiescher, M., & Käppeler, F. 2014, ApJ,

787, 10Bisterzo, S., Travaglio, C., Wiescher, M., Käppeler, F., & Gallino, R. 2017, ApJ,

835, 97Bragaglia, A., & Tosi, M. 2006, AJ, 131, 1544Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127Burbidge, E. M., Burbidge, G. R., Fowler, W. A., & Hoyle, F. 1957, Reviews of

Modern Physics, 29, 547Busso, M., Gallino, R., & Wasserburg, G. J. 1999, ARA&A, 37, 239Busso, M., Wasserburg, G. J., Nollett, K. M., & Calandra, A. 2007, ApJ, 671,

802Cantat-Gaudin, T., Vallenari, A., Zaggia, S., et al. 2014, A&A, 569, AA17Carraro, G., Geisler, D., Moitinho, A., Baume, G., & Vázquez, R. A. 2005, A&A,

442, 917Carraro, G., Janes, K. A., Costa, E., & Méndez, R. A. 2006, MNRAS, 368, 1078Cignoni, M., Beccari, G., Bragaglia, A., & Tosi, M. 2011, MNRAS, 416, 1077Clem, J. L., Landolt, A. U., Hoard, D. W., & Wachter, S. 2011, AJ, 141, 115Côté, B., Fryer, C. L., Belczynski, K., et al. 2017, arXiv:1710.05875Cowan J. J., Rose W. K., 1977, ApJ, 212, 149Cristallo, S., Straniero, O., Gallino, R., et al. 2009, ApJ, 696, 797da Silva, R., Porto de Mello, G. F., Milone, A. C., et al. 2012, A&A, 542, A84Delgado Mena, E., Tsantaki, M., Adibekyan, V. Z., et al. 2017, A&A, 606, A94Dell’Omodarme, M., Valle, G., Degl’Innocenti, S., & Prada Moroni, P. G. 2012,

A&A, 540, A26Denissenkov, P. A., Herwig, F., Truran, J. W., & Paxton, B. 2013, ApJ, 772, 37Dias, W. S., Alessi, B. S., Moitinho, A., & Lépine, J. R. D. 2002, A&A, 389, 871Donati, P., Bragaglia, A., Cignoni, M., Cocozza, G., & Tosi, M. 2012, MNRAS,

424, 1132Donati, P., Cantat Gaudin, T., Bragaglia, A., et al. 2014b, A&A, 561, AA94D’Orazi, V., Magrini, L., Randich, S., et al. 2009, ApJ, 693, L31Feltzing, S., Howes, L. M., McMillan, P. J., & Stonkute, E. 2017, MNRAS, 465,

L109Friel, E. D., Donati, P., Bragaglia, A., et al. 2014, A&A, 563, AA117Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al. 2016, A&A, 595, A1Gallino, R., Arlandini, C., Busso, M., et al. 1998, ApJ, 497, 388Gilmore, G., Randich, S., Asplund, M., et al. 2012, The Messenger, 147, 25González Hernández, J. I., Israelian, G., Santos, N. C., et al. 2010, ApJ, 720,

1592Grevesse, N., Asplund, M., & Sauval, A. J. 2007, Space Sci. Rev., 130, 105Herwig, F. 2000, A&A, 360, 952Karakas, A. I., & Lugaro, M. 2016, ApJ, 825, 26Kordopatis, G., Recio-Blanco, A., de Laverny, P., et al. 2011, A&A, 535, A107Kordopatis, G., Wyse, R. F. G., Gilmore, G., et al. 2015b, A&A, 582, A122Jacobson, H. R., & Friel, E. D. 2013, AJ, 145, 107Jacobson, H. R., Friel, E. D., & Pilachowski, C. A. 2011, AJ, 141, 58Jacobson, H. R., Friel, E. D., Jilkova, L., et al. 2016, A&A, 591, 37Liu, F., Asplund, M., Yong, D., et al. 2016, MNRAS, 463, 696Maiorca, E., Randich, S., Busso, M., Magrini, L., & Palmerini, S. 2011, ApJ,

736, 120Maiorca, E., Magrini, L., Busso, M., et al. 2012, ApJ, 747, 53Magrini, L., Sestito, P., Randich, S., & Galli, D. 2009, A&A, 494, 95Magrini, L., Randich, S., Donati, P., et al. 2015, A&A, 580, A85Magrini, L., Randich, S., Kordopatis, G., et al. 2017, A&A, 603, A2Mashonkina, L., & Gehren, T. 2001, A&A, 376, 232Matteucci, F., Romano, D., Arcones, A., Korobkin, O., & Rosswog, S. 2014,

MNRAS, 438, 2177Mermilliod, J.-C., Clariá, J. J., Andersen, J., Piatti, A. E., & Mayor, M. 2001,

A&A, 375, 30Mishenina, T., Korotin, S., Carraro, G., Kovtyukh, V. V., & Yegorova, I. A. 2013,

MNRAS, 433, 1436Mishenina, T., Pignatari, M., Carraro, G., et al. 2015, MNRAS, 446, 3651Modigliani, A., Mulas, G., Porceddu, I., et al. 2004, The Messenger, 118, 8Nissen, P. E. 2016, A&A, 593, A65Nissen, P. E., Silva Aguirre, V., Christensen-Dalsgaard, J., et al. 2017,

arXiv:1710.03544Nordhaus, J., Busso, M., Wasserburg, G. J., Blackman, E. G., & Palmerini, S.

2008, ApJ, 684, L29Önehag, A., Gustafsson, B., & Korn, A. 2014, A&A, 562, A102Overbeek, J. C., Friel, E. D., & Jacobson, H. R. 2016, ApJ, 824, 75Pagel, B. E. J., & Tautvaisiene, G. 1997, MNRAS, 288, 108Palmerini, S., Cristallo, S., Busso, M., et al. 2011, ApJ, 741, 26Pancino, E., Lardo, C., Altavilla, G., et al. 2017, A&A, 598, A5Pasquini, L., Avila, G., Blecha, A., et al. 2002, The Messenger, 110, 1Pasquini, L., Biazzo, K., Bonifacio, P., Randich, S., & Bedin, L. R. 2008, A&A,

489, 677Piatti, A. E., Clariá, J. J., Bica, E., Geisler, D., & Minniti, D. 1998, AJ, 116, 801

Pignatari, M., Gallino, R., Heil, M., et al. 2010, ApJ, 710, 1557

Raiteri, C. M., Gallino, R., Busso, M., Neuberger, D., & Kaeppeler, F. 1993, ApJ,419, 207

Raiteri, C. M., Villata, M., Gallino, R., Busso, M., & Cravanzola, A. 1999, ApJ,518, L91

Randich, S., Sestito, P., Primas, F., Pallavicini, R., & Pasquini, L. 2006, A&A,450, 557

Randich, S., Gilmore, G., & Gaia-ESO Consortium 2013, The Messenger, 154,47

Randich, S., Tognelli, E., Jackson, R., et al. 2017, arXiv:1711.07699

Recio-Blanco, A., de Laverny, P., Kordopatis, G., et al. 2014, A&A, 567, A5

Reddy, A. B. S., & Lambert, D. L. 2017, ApJ, 845, 151

Sacco, G. G., Morbidelli, L., Franciosini, E., et al. 2014, A&A, 565, A113

Salaris, M., Weiss, A., & Percival, S. M. 2004, A&A, 414, 163

Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525

Seeger, P. A., Fowler, W. A., & Clayton, D. D. 1965, ApJS, 11, 121

Sharma, S., Pandey, A. K., Ogura, K., et al. 2006, AJ, 132, 1669

Silva Aguirre, V., Lund, M. N., Antia, H. M., et al. 2017, ApJ, 835, 173

Smiljanic, R., Korn, A. J., Bergemann, M., et al. 2014, A&A, 570, A122

Sneden, C., Cowan, J. J., & Gallino, R. 2008, ARA&A, 46, 241

Spina, L., Palla, F., Randich, S., et al. 2015, A&A, 582, L6

Spina, L., Meléndez, J., Karakas, A. I., et al. 2016, A&A, 593, A125

Spina, L., Meléndez, J., Karakas, A. I., et al. 2018, MNRAS, 474, 2580

Stonkute, E., Koposov, S. E., Howes, L. M., et al. 2016, MNRAS, 460, 1131

Thielemann, F.-K., Arcones, A., Käppeli, R., et al. 2011, Progress in Particle andNuclear Physics, 66, 346

Tognelli, E., Prada Moroni, P. G., & Degl’Innocenti, S. 2011, A&A, 533, A109

Travaglio, C., Galli, D., Gallino, R., et al. 1999, ApJ, 521, 691

Travaglio, C., Gallino, R., Arnone, E., et al. 2004, ApJ, 601, 864

Trippella, O., Busso, M., Palmerini, S., Maiorca, E., & Nucci, M. C. 2016, ApJ,818, 125

Trippella, O., & La Cognata, M. 2017, ApJ, 837, 41

Yong, D., Carney, B. W., & Friel, E. D. 2012, AJ, 144, 95

Zhao, G., Mashonkina, L., Yan, H. L., et al. 2016, ApJ, 833, 225


http://arxiv.org/abs/1310.4578





1 INAF - Osservatorio Astrofisico di Arcetri, Largo E. Fermi, 5, I-50125 Firenze, Italy e-mail: [email protected]

2 Monash Centre for Astrophysics, School of Physics and Astronomy,Monash University, VIC 3800, Australia

3 Department of Astronomy, Indiana University, Bloomington, IN,USA

4 Université Côte d’Azur, Observatoire de la Côte d’Azur, CNRS,Laboratoire Lagrange, France

5 Institute of Astronomy, Madingley Road, University of Cambridge,CB3 0HA, UK

6 Space Science Data Center - Agenzia Spaziale Italiana, Via del Po-litecnico SNC, 00133 Roma

7 INAF-Osservatorio di Astrofisica e Scienza dello Spazio, via Gob-etti 93/3, I-40129 Bologna, Italy

8 Astronomical Observatory, Institute of Theoretical Physics and As-tronomy, Vilnius University, Sauletekio al. 3, LT-10222 Vilnius,Lithuania

9 Instituto de Astrofisica e ciencias do espaço -CAUP, Universidadedo Porto, Rua das Estrelas, P-4150-762 Porto, Portugal

10 Departamento de Astrofísica, Centro de Astrobiología (INTA-CSIC), ESAC Campus, Camino Bajo del Castillo s/n, E-28692 Vil-lanueva de la Cañada, Madrid, Spain

11 INAF - Osservatorio Astronomico di Palermo, Piazza del Parla-mento 1, 90134, Palermo, Italy

12 European Southern Observatory, Alonso de Cordova 3107 Vitacura,Santiago de Chile, Chile

13 Departamento de Ciencias Fisicas, Universidad Andres Bello, Fer-nandez Concha 700, Las Condes, Santiago, Chile

14 Lund Observatory, Department of Astronomy and TheoreticalPhysics, Box 43, SE-221 00 Lund, Sweden

15 Astrophysics Group, Keele University, Keele, Staffordshire ST55BG, United Kingdom

16 INAF - Padova Observatory, Vicolo dell’Osservatorio 5, 35122Padova, Italy

17 Instituto de Astrofísica de Andalucía-CSIC, Apdo. 3004, 18080,Granada, Spain

18 GEPI, Observatoire de Paris, CNRS, Université Paris Diderot, 5Place Jules Janssen, 92190 Meudon, France

19 Department of Physics and Astronomy, Uppsala University, Box516, SE-751 20 Uppsala, Sweden

20 Laboratoire Lagrange (UMR7293), Université de Nice Sophia An-tipolis, CNRS,Observatoire de la Côte d’Azur, CS 34229,F-06304Nice cedex 4, France

21 Nicolaus Copernicus Astronomical Center, Polish Academy of Sci-ences, ul. Bartycka 18, 00-716, Warsaw, Poland

22 Instituto de Física y Astronomiía, Universidad de Valparaiíso, Chile23 Millennium Nucleus “Núcleo Planet Formation", Universidad de

Valparaíso24 Dipartimento di Fisica e Astronomia, Università di Padova, Vicolo

dell’Osservatorio 3, 35122 Padova, Italy25 Departamento de Didáctica, Universidad de Cádiz, 11519 Puerto

Real, Cádiz, Spain26 INAF - Osservatorio Astrofisico di Catania, via S. Sofia 78, 95123,

Catania, Italy27 Núcleo de Astronomía, Facultad de Ingeniería, Universidad Diego

Portales, Av. Ejército 441, Santiago, Chile28 Instituto de Astrofísica de Canarias, E-38205 La Laguna, Tenerife,

Spain29 Universidad de La Laguna, Dept. Astrofísica, E-38206 La Laguna,

Tenerife, Spain


Documents

The Gaia-ESO Survey: the origin and evolution of s-process elements · 2018. 6. 11. · and Zr, and three belonging to the second peak, Ba, La and Ce - using the Gaia-ESO idr5 results